Restriction Orifice with High DP and % of liquid upstream 3

[Steve010]

I have a problem in a project i am working in now (still in the engineering phase).
The purge line for the flare is from the discharge of a compressor. so the design was to place a restriction orifice to drop the pressure from 1500 psi to 20 psi. I calculated the bore diameter of the RO and it was found to be 4 mm. The temperature downstream of the RO will drop to -35 C. The process engineer said that hydrates will form so decided to inject methanol upstream of the RO to prevent hydrate formation. The problem is that the required methanol injection for hydrate prevention made the stream a 2 phase flow (gas +2% methanol). I cannot place a drain hole because the drain hole should not exceed 1/10 of the RO bore which will yield a a drain hole diameter of 0.4 mm which is not practical. Also if i used a standard drain hole of 2 mm the flow will increase dramatically (because the drain hole diameter will be 1/2 the bore diameter). I am concerned about the accumulation of the liquid (methanol) at the RO upstream face. Note that this is the only source of purge for the flare and if it got plugged a flare flashback might occur.

Any ideas?

 

[apetri]

for hydrates you must have water,
and if you need to inject 2% molar fraction of methanol you have something more than 0.000..1% of water (supposing values are correct), that means that other problems (in addition to hydrate formation) are possible...
Anyway, you can install the RO horizontal (flow up->down), use a concave plate etc. etc. with a restriction orifice you are free to adopt a variant...
A difficult frequently underestimated is related with evaluation of RO area,
in case of doubts (supercritical, two phases etc.) better to calculate the speed of sound with a EOS or similar methods,
for two phases HEM model should give reasonable values,
you may find additional information in this thread
"

http://www.cheresources.com/invisio...-sound-in-vapor-liquid-hydrocarbons-mixture/"

 

[Steve010]

Thanks for your response.
Do you mean by concave orifice: that the bore will be concave, or the whole restriction orifice plate will be concave ? I only know the concave bores and I think methanol drainage will be unpredictable even with flow from top to bottom. If there exists a restriction orifice plate with a concave body the methanol drainage will be guaranteed. Regarding the calculation i think that if i calculate the bore using a single phase gas flow the error will not be that much(methanol is 2% by volume). My concern is that for the gas (with 2% methanol)chocked flow conditions the methanol droplets will reach sonic velocity and might influence the downstream piping. What do you think ?

Also is it normal to take the flare purge line from such location (compressor discharge) which yielded in all these problems?

 

[Steve010]

What are the other problems (in addition to hydrate formation)?

 

[apetri]

you are mainly concerned with draining of some liquid fraction
but from a process point of view if there is a relatively large amount of water
(presuming it's in vapor state after compression) and this steam/water goes below 0 C (freezing point, without methanol addition) due to expansion, I would be more concerned about solving correctly phase equilibria...
Anyway, if you wish to purchase the RO sure the manufacturer can help with some specific design as well as materials suitable for your application...
About flare purge line, no, it's not usual but you may ask the reason to the process engineer who does the calc's, we can't help without detailed information...

 

[EmmanuelTop]

Purge flow is usually very low, so the true temperatures will likely be considerably above the calculated expansion temperature. The solution with injecting MeOH doesn't sound like the best approach anyway.

Did you consider using compressor suction gas or plant N2 for purge purposes? It would help you get rid of all the problems mentioned in your OP.


Dejan IVANOVIC
Process Engineer, MSChE

 

[apetri]

>Purge flow is usually very low
I agree, presuming it's natural gas with that RO flow should be below 1000 Kg/h ,
may be they can consider alternatives...

 

[Steve010]

Yes i tried to convince the process engineer to take from the compressor suction but he refused. His only reason was that the compressor discharge is unintermittent, because it is connected to the discharges of many other compressors (from different projects), so that if one compressor shuts down the flow will still be maintained.
Yes the flow is low 500 kg/hr and he calculated the RO downstream temperature using these conditions.
I am an instrumentation and control engineer not a process engineer but i also agree that MeOH injection isn't the best approach as i do not believe that continuous MeOH injection could be easily maintained as per the injection rate he calculated.
Also i am concerned about Meoh droplets exiting the RO at sonic velocity (might influence the downstream piping)

 

[Steve010]

I believe that with just 2% by volume MeOH, i can assume single phase gas flow for the RO bore sizing. Am i correct ?

 

[EmmanuelTop]

Since the process engineer got you in this mess in the first place, what is his approach in the design of this RO and the resulting consequences? Or has he come with the idea and left you do do the work?

It is hard to believe that there are no better sources of fuel gas. What drives the compressors? Is there a fuel gas source on site? Nitrogen? What about the common suction header from where all compressors are fed?

Dejan IVANOVIC
Process Engineer, MSChE

 

[Steve010]

There are many other sources that can be used as the purge line fuel gas. The pilot gas for the flare is coming from the fuel gas scrubber and it is very simple to take purge gas from the same scrubber (i think).
I am just trying to solve the problem from an instrument perspective, so that i have a backup plan if he continued on insisting on the compressor discharge location.

Any ideas ?

 

[EmmanuelTop]

Can you take the hot discharge (upstream of the compressor discharge cooler) and expand a slipstream across the orifice? See what temperature you get in that case. It may be too hot but for sure Methanol injection would not be required.

In my opinion, everything is better than what has been proposed originally - continuous injection of MeOH in the flare header and a two-phase orifice. It is never going to work as intended.

Dejan IVANOVIC
Process Engineer, MSChE

 

[LittleInch]

A couple of things strike me when looking at this.

1) Why are you taking such a huge pressure drop in one go? The noise from this will be horrendous - has anyone worked this out? If you're going down this route, which seems to be quite silly to me, then you need to do it in three or four steps minimum and allow some pipework between the stages so that you can get some ambient heat into the stages.

2) Why not use a proper pressure let down skid with small electric heater - basically a small fuel gas skid?

3) A 4mm RO is just ridiculous for constant flow - it will either fairly rapidly become an 8mm (or bigger) RO or get blocked with something very easily, especially during commissioning or maintenance.

40 You need to run this past a more experienced engineer who understands the practicality of life, not just that the calculation work out correctly.

This needs to go back to the concept, not trying to fix what was always a bad idea.

Good luck and let us know how you get on.

sorry for being so negative, but this looks like a triumph for spread sheet design over real life.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[casflo]

To avoid the holding back of the liquid in the RO plate use a multi-hole RO plate placed in a vertical position.
casflo

 

[Steve010]

Littleinch. I agree with every single word you said (and that is what i am trying to convince the process engineer with)except for the noise. i calculated the noise and it was 89 dB. Thank you

 

[LittleInch]

89db is still pretty high and will resonate down a flare line.... No one, but no one will ever thank him for installing this thing and I doubt it would ever work properly.

Good luck.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Use the low pressure fuel gas.
The other problem with 1500psig gas dropping to 20psig is the brittle fracture concern for the downstream piping - at -35degC, you cannot use A106B, it will have to be A333 Gr P6 or similar. The risk of loss of containment due to line rupture might give this process engineer the shivers ?

 

[Steve010]

What do you think of using a pressure control loop with a single pressure control valve that reduces the pressure to 20 psig ?
Is such an installation reliable (provided that the flow at valve fail open will not make the downstream pressure exceed the design limits of the flare knock out drum) ?
I know the control valve bore diameter will be small as well, but i also know that control valves can be used for two phase flow.

 

[georgeverghese]

Dont see why you would want to use a PIC / PCV loop on purge gas ? The RO should work ok.

At -35degC, the other concern is with solid water ice forming on the inner pipe walls.

 

[Steve010]

Because there is methanol liquid entrainment upstream of the RO and the RO bore (4 mm) is very small. Liquid methanol can accumulate upstream of the RO